Otter.ai adds AI-generated meeting summaries and new home feed
Source: https://www.theverge.com/2022/3/29/22999527/otter-ai-transcription-meeting-summary-news-feed-feature

Otter.ai made its name as a slick, AI-powered transcription service. But as this type of automated transcription becomes more commonplace, the company is expanding its remit — adding a host of features, including AI-generated meeting summaries, with the aim of turning users’ Otter accounts into collaborative hubs for work.

The goal is to make Otter bigger than transcriptions and cater to the company’s growing number of enterprise customers. “A year ago, most of [our customers] were individuals, but more and more professionals are using it,” Otter.ai CEO Sam Liang told The Verge in an interview. “The new Otter makes it a one-stop for all your meeting contents and collaboration needs.”

When accessing accounts on the web, Otter users will now see a “home feed” that pulls together transcriptions and a calendar of upcoming meetings into a single overview. They’ll be able to jump into meetings directly from their calendar and use Otter integrations with services like Zoom, Microsoft Teams, and Google Meet to record and transcribe the audio.

The transcriptions can then be added to in various ways. The big new feature is AI-generated meeting summaries, which are supposed to highlight the most important moments in recordings. There’s also something Otter is calling “meeting gems.” These are parts of the transcript that have been highlighted by users, who can then tag in co-workers and add comments or tasks. Users can also now add screenshots to transcripts with a single click, making it easier to reference visual material discussed during meetings.

The most intriguing feature, though, is AI-generated meeting summaries. We haven’t been able to test this for ourselves, though even Liang admitted the tool was “far from perfect, but it’s a great start.”

The company’s software looks at a lot of different factors to decide what are the most relevant points from a meeting, says Liang. “We look at the topic words people use. We look at the speaker dynamics — who is talking and what topics they discuss [...] and when did they change topic. It’s never just based on one signal — it’s always a combination.”

In a preview of the software we were shown via Zoom, the tool seemed to pick out when new topics of conversation were introduced and the speaker changed. It could potentially be a useful way to skip through relevant parts of a meeting, but it’s very unlikely the machine learning could match the knowledge of a human, who would know far more about the background and context of a meeting and its participants.

In addition to the summaries, Otter also offers a breakdown of who spends the most time in a meeting talking — a tool that could be useful when trying to balance collaboration in teams. Liang says there’s much more analysis that could be done, too (like sentiment analysis on the language used) that would let Otter expand far beyond its current space. “This is why I say Otter, potentially, has a bigger total addressable market than Zoom or other conferencing systems,” he says. “The conferencing systems just provide a way for people to talk to each other; they don’t really understand what they’re discussing.”

Other startups are already moving fast in this space, though. One called Poised promises to coach users on their presentation skills by transcribing meetings and analyzing things like their use of filler words and speaking speed. Another called Sembly offers similar AI-generated meeting summaries.

For Otter, though, the bigger threat is from juggernauts like Google and Microsoft, whose AI expertise would allow them to quickly create such features themselves and offer them to a far larger audience. (Indeed, they’re already ahead. Examples include Microsoft’s PowerPoint, which offers its own speech analysis and tips for presenters, and Google Docs, which uses AI to generate summaries and content pages.) When asked about this threat, Liang says that Otter will succeed for the reason so many startups do: it’s focused on a single product while tech giants are distracted by their sprawling interests.

“The question is: how obsessed are you?” he says. “Eric Yuan, the Zoom CEO, is, I bet, way more obsessed with video quality than the Google CEO. The Google CEO makes 99 percent of their money from Search and YouTube, so nothing else matters.”

For Otter, he says, that obsession is turning meeting transcripts into action plans. Now, the company itself has to follow through.



Source: https://www.theverge.com/2022/3/29/22999527/otter-ai-transcription-meeting-summary-news-feed-feature

Leave a Reply

Your email address will not be published. Required fields are marked *